论文部分内容阅读
为了降低传输失真对自由视点视频虚拟视点质量的影响,提出了一种同时适用于彩色图及深度图的宏块区分模型。模型主要包括两部分:首先结合立体视频时空域的相关性,在考虑错误扩散的基础上,提出一个递归形式的宏块级传输失真模型;之后分别讨论彩色图失真和深度图失真对虚拟视点质量的影响,进而提出了一个低复杂度的基于绘制失真分析的重要性模型。实验结果表明,与随机丢包相比,本文算法能在不改变丢包率(PLR)的情况下大幅提高虚拟视点的客观质量,在PLR为20%时,平均峰值信噪比(PSNR)最大能提高15.65dB,且主观质量接近零传输失真情况。
In order to reduce the impact of transmission distortion on the quality of virtual view point of free-view video, a macroblock distinguishing model suitable for both color image and depth map is proposed. The model mainly consists of two parts: Firstly, a recursive form of macroblock-based transmission distortion model is proposed based on the correlation of the spatial and temporal regions of stereo video, and the error diffusion is considered. Then the effects of color and distortion distortion on the virtual viewpoint quality Then, a low-complexity model of importance based on the analysis of distortions is proposed. The experimental results show that compared with random packet loss, the proposed algorithm can significantly improve the objective quality of virtual viewpoints without changing the PLR. When the PLR is 20%, the average peak signal-to-noise ratio (PSNR) is the largest Can improve 15.65dB, and subjective quality close to zero transmission distortion.